Acorn application guide

Filth Finder

Use text matches to prioritise a lawful examination, not to classify unseen image content. The demonstrated workflow makes no visual-content verdict.

Demonstrated on constructed training data

Do filenames or stored text contain leads requiring a properly authorised content review?

Start with

Constructed filenames, metadata and database text used for keyword review.

What you can take away

Keyword and metadata leads from paths and database text.

The native scanner reports three filename keyword hits and two database string hits, with CSV exports for review. Content hashing and skin-tone scoring are visibly disabled.

Open the screenshot and choose “View actual size” to read the records at their original resolution.

A practical starting point

How the workflow fits together.

Agree the review scope and relevant search terms.
Inspect path and database-text matches with their context.
Refer any relevant material for appropriate examination rather than treating a keyword as a classification.

Useful next steps

Start with your requirements

Bring a sample question.

Ask us to demonstrate this workflow with suitable public or constructed material. Do not send confidential evidence in an initial enquiry.

Planned for Q1 2027

Be first to hear. Save 10% at launch.

Request Acorn launch news and details of the planned 10% offer. No deposit or purchase commitment.

Notify me at launch ↗

Launch-update requests go to the SQFR team for review. Final offer terms and product availability are still to be confirmed.

About the screenshots and illustrations

Application screens are selected from the September 2026 Acorn screenshot pack. Captions distinguish native setup views, constructed training records and public-corpus results. They are not private client cases, and a displayed control does not establish that every operation was completed.

Relevant public sources include DeepBlueCLI training event logs and Plaso test data. Check the relevant source terms before redistributing an underlying dataset.

Workspace scenes and sector mascot variants are generated illustrations. They do not show actual police, judicial, military or university deployments or endorsements. The original Squirrel Forensics identity is retained.